-
- Up one level
- Correlated Data - David Kleinbaum
These videos contain recordings of lectures given by Dr. David G. Kleinbaum in his 2016 course on the Analysis of Correlated Data given at Emory University. Topics include an overview of modeling issues for correlated data (including notation, data layout, goals of analysis), matrix formulation of correlated data models, linear models for analyzing correlated data, SAS's Mixed procedure, non-linear models for analyzing correlated data, GEE estimation using SAS's GENMOD and GLIMMIX procedures, tests for correlation structures using the COVTEST statement in GLIMMIX, choice of correlation structure, and empirical vs. model-based estimates of variance. (Note: Jan 26 class taught by Zach Binney)
- Improving Medical Statistics and the Interpretation of Medical Studies
Featured in NetWatch, Science, 6 May 2005
Examples of the misuse of statistics and inappropriate conclusions in the medical literature, from Eric Roehm, M.D., F.A.C.C. a practicing cardiologist with a long-standing interest in the proper interpretation of medical studies. Roehm created this web site in 2005, retired from private practice as of 2007, and remained involved with multiple nonprofit projects and websites in the medical field.
- Interpreting Cohen's d effect size an interactive visualization
Created by Kristoffer Magnusson. Gain insight into comparing Gaussian ("normal") distributions
- Intro to Biostatistics; Distributions and Inference - David Kleinbaum
The videos represent narrated PowerPoint presentations developed by Dr. David G. Kleinbaum on important topics covered in introductory biostatistics courses for non-statistics majors in public health or related fields. The videos were originally constructed in the 1970's as slide-tape presentations by Dr. Kleinbaum but were converted to PowerPoint presentations in 2012. This collection of 8 videos represents material often covered in the first half of a biostatistics course, so other material (e.g., comparing two parameters, chi-square tests, linear regression) is not included here. Viewing study guides corresponding to each video presented here can be obtained by clicking on the links provided below the description of the video underneath the video title. These study guides summarize the topic being covered, provide a worked example involving data, and give a post-test (with answers provided) on the subject matter.
- Measurement error and the replication crisis
Eric Loken, Andrew Gelman. Science 10 Feb 2017;355(6325):584-585 Measurement error adds noise to predictions, increases uncertainty in parameter estimates, and makes it more difficult to discover new phenomena or to distinguish among competing theories. A common view is that any study finding an effect under noisy conditions provides evidence that the underlying effect is particularly strong and robust. Yet, statistical significance conveys very little information when measurements are noisy. In noisy research settings, poor measurement can contribute to exaggerated estimates of effect size. This problem and related misunderstandings are key components in a feedback loop that perpetuates the replication crisis in science.
- See epidemiolog.net for additional resources and tools
- VassarStats: Website for Statistical Computation - Richard Lowry
VassarStats is a user-friendly tool for performing statistical computation. Each of the links in the left-hand panel displays an annotated list of the statistical procedures available under that rubric. A «Site Map» has a complete list of all available items. There is also a companion online textbook linked on the homepage.